GPT Taste — Awwwards-Level Frontend Design Skill for Muse
Build company websites at the level of figma.com and x.ai — plain copy, real photographs, one section per screen.
- What
- Build company websites at the level of figma.com and x.ai — plain copy, real photographs, one section per screen.
- Cost
- Free
- Needs
- Node 22+ and Chrome, Edge or Chromium (ffmpeg optional). No external services, APIs, or accounts. Output quality depends on your brief and your real material (logo, screens, photographs, prices, address, hours); the checkers need a real browser to run.
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
tasteskill 2.5.0, the anti-slop frontend skill: builds professional company websites at the level of figma.com and x.ai — plain copy, one system, one section per screen, equal edges, real photographs and real product UI (a product render when there are no photographs), with shader effects, scenes and moments only on tech, gaming and software pages, quiet micro-interactions, and a checker that enforces it in a real browser (consistency.mjs, check.mjs, perf.mjs). It deals the hero, palettes and pairings instead of choosing by taste (deal.mjs, start.mjs, Node's Math.random() over the kind's lists), follows 9 house rules that win over everything else, and needs Node 22+ and Chrome, Edge or Chromium. Use it whenever you build, design, redesign, restyle, polish or review a landing page, company, product, shop, cafe, hotel, studio, SaaS, AI, developer-tool or game site. Discovered via skills.sh. Honest note: this is a skill plus a starter kit and checkers, not a component library — output quality still depends on your brief and your real material (logo, screens, photographs, prices, address, hours), and the checkers need a real browser to run. Skill Harbor never reviews the code, review it yourself before use. Not verified.
Curated by Skill Harbor — free, open-source skill (MIT) by @leonxlnx.
Version:
Quick answer
GPT Taste is a free Muse skill auditing any design against GPT's interface taste, turning Muse into an elite frontend engineer. It aggressively breaks LLM defaults, using simulated randomization of hero architecture, typography and layout choices to escape generic looks. The goal is Awwwards-level frontend design: distinctive, confident interfaces that feel designed rather than generated from the same tired templates.
Install
Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Nov 11, 2026📌 Pinned version: commit f578e23Step 0 — before installing anything: read this install prompt, summarize in a few lines what this build does and what installing it involves, then ask me to confirm before you proceed. Prerequisites: Node 22+ and Chrome, Edge or Chromium (ffmpeg optional). No external services, APIs, or accounts. Output quality depends on your brief and your real material (logo, screens, photographs, prices, address, hours); the checkers need a real browser to run. Install "tasteskill 2.5.0 — the anti-slop frontend skill for Muse" (upstream skill name design-taste-frontend) for me. It builds professional company websites at the level of figma.com and x.ai: plain copy, one system, one section per screen, equal edges, real photographs and real product UI (a product render when there are no photographs), with shader effects, scenes and moments only on tech, gaming and software pages, quiet micro-interactions, and a checker that enforces it in a real browser. Use it whenever you build, design, redesign, restyle, polish or review a landing page, company, product, shop, cafe, hotel, studio, SaaS, AI, developer-tool or game site, in any stack (HTML/CSS, React, Next.js, Vue, Svelte, Astro, Tailwind), or when the user says a page looks generic, templated or "AI-made". How it varies pages: it deals the hero, colour palettes and type pairings instead of choosing by taste — deal.mjs with a history file, start.mjs deal/css/brand, and a Node one-liner using Math.random() over the kind's lists in references/sections.md — so no two pages look alike. Its 9 house rules win over everything else (no caption under any visual, every visual renders, one section per screen, equal edges, hover micro-animations, more and better real photographs, real full-colour icons, award-site scenes and moments on tech only, privacy-safe addresses). Motion comes from the skill's own CSS motion layer (assets/motion) and shader/WebGL effects on tech pages only. Repository: https://github.com/leonxlnx/taste-skill Pinned version: install the code at commit f578e23218be09ffd5b70debbfdec9fe8286d310 exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. Updated November 8, 2026: Skill Harbor regenerated this install prompt from the current upstream file (tasteskill 2.5.0, skills/taste-skill/SKILL.md). What changed since the version the previous prompt described: the previous prompt described a different skill from the same repository, not this file. This prompt now describes the file itself: it deals hero layouts, colour palettes and type pairings with a history file (deal.mjs, start.mjs, Node's Math.random()), enforces 9 house rules, needs Node 22+ and Chrome, Edge or Chromium, and checks the built page in a real browser with consistency.mjs, check.mjs and perf.mjs, with photo credits from photos.mjs. The install folder is now "taste-skill" (the upstream directory), not "gpt-taste". The full list of the removed misdescriptions is in the regen note that accompanies this prompt. The integrity fingerprint below was updated to the new file. 1. Fetch the SKILL.md file for the "taste-skill" skill (skills/taste-skill/SKILL.md, and any helper files, including the assets/, library/, references/ and scripts/ folders) from the repository into a temporary folder 2. Integrity check: compute the SHA-256 of the fetched SKILL.md; it must match f2fe0fb10f4c7d742866d1869a2ce4b16d14e9819612bf0245738705d806062f, the fingerprint Skill Harbor recorded on November 8, 2026. If it does not match, STOP, do not install, and tell me, showing the expected and actual hashes. 3. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code, credentials only via the secure vault. Verify that holds here; STOP on any red flag and tell me. 4. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "taste-skill". 5. Verify with no network calls: frontmatter valid, files in place. 6. Report what was installed, where, and what I still need to do myself (e.g. confirm my brief — what is sold, who buys it, the one main action, and the real material: logo, screens, photographs, prices, address, hours — so the skill can build the page; confirm Node 22+ and Chrome, Edge or Chromium are available for the checker). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Saved to your recent installs. Find it anytime on /connect.
🔒 Security, provenance and what this touches
The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?
Verification certificate
Dated facts about what was checked for this listing, and how. This is not a safety certification.
- Pinned version
- f578e23 · 2026-10-13
- Sandbox install test
- Install ran in our test sandbox · Nov 11, 2026
- Creator identity
- Not verified
- Fingerprint watch
- Last checked · Nov 19, 2026 : no change detected in the files linked by this install prompt
- Install prompt analysis
- 2026-10-13 : No commands, network calls, packages, file writes or API keys detected.
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Questions
How do I install a skill?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.